Authority in Visual Exegesis
Jeffrey Moser
Studies of illustrated texts generally proceed from the assumption that the illustrations are subordinate to the text in the production of meaning. Although we recognise that illustrations fundamentally transform the experience of a book, as a practical matter, our inquiries typically start from the words. This is as true for inscribed pictures as it is for illustrated texts. Just as the presence of cartouches naming the emperors in the Thirteen Emperors handscroll attributed to Yan Liben (d. 673), encourages scholars to seek correspondences between the visual qualities of the portraits and the historical biographies of their named subjects,1 the imagery on the frontispiece of the twelfth-century Floreffe Bible is perceived as “anomalous” because it contradicts the iconographic expectations generated by its written tituli.2 Despite the purportedly visual focus of art historical inquiry, wherever the written word is seen, it consistently instigates the leading questions and guides their resolution, regardless of whether the subject is an early Chinese handscroll or a medieval European frontispiece.
The agency of words is doubly evident when images are related to canonical texts like the Bible. In an overview of the relationship between art and biblical exegesis in the Romanesque and Gothic periods, Christopher Hughes identifies three categories of exegetical art: illustrations in exegetical texts, art illustrating exegetical writing and thought, and art as visual exegesis.3 In the simplest terms, the distinction between these three categories is prepositional: art in biblical interpretation, art of biblical interpretation, and art as biblical interpretation. What is most telling for our purposes is that even the third category, which gives the greatest agency to the visual by considering pictures in themselves as a form of exegesis, the presumption of a stable base text persists. Even when its visual character is foregrounded, the classification
Jeffrey Moser is Assistant Professor of East Asian Art History at McGill University.
1
Chen Baozhen, ‘Painting as History: A Study of the Thirteen Emperors attributed to Yan Liben,’ in The History of Painting in East Asia: Essays on Scholarly Method (Taipei: Rock Publishing International, 2008), pp. 55-92.
2
Anne-Marie Bouché, ‘Vox Imaginis: Anomaly and Enigma in Romanesque Art,’ in The Mind’s Eye: Art and Theological Argument in the Middle Ages, eds Jeffrey Hamburger and Anne-Marie Bouché (Princeton: Princeton University Press, 2006), pp. 306-335.
3
Christopher Hughes, ‘Art and Exegesis,’ in A Companion to Medieval Art: Romanesque and Gothic in Northern Europe, ed. Conrad Rudolph (Blackwell Publishing, 2006), pp. 180-187.
of this art as exegetic necessarily makes it belated; we interpret it to understand not the Bible but a biblical hermeneutic. In the end, it is the presumed stability of the Bible that makes the specification of this hermeneutic as a historical entity possible. Art as exegesis preserves the paradigmatic primacy of the text.
This textual authority is not unique to the elaborate and varied hermeneutics developed by the West Asian and European religions of the book. The idea of foundational texts with allegorical or other hidden meanings similarly motivates the long East Asian tradition of exegesis on the Confucian classics. Stephen Owen observes that concern for “that which the text conceals” has guided Chinese classical exegesis from the beginning, and that although the Chinese tradition was largely disinterested in Platonic absolutes, it recognised a surface-substance distinction consistent with the Greek ‘disclosedness’ (aletheia) as a justification for exegetic inquiry.4
The influence of the canonical text on art historical interpretation is plainly evident in the best-known studies of Chinese exegetic art, such as Richard Barnhart’s influential examination of Li Gonglin’s Illustrations of the
Classic of Filial Piety,5 and Julia Murray’s seminal study of the various
Illustrations of the Book of Odes attributed to the Southern Song court artist
Ma Hezhi.6 Both scholars examine the ways in which the artists respond to, elaborate upon, and otherwise translate the relevant classic into graphic form. Although they recognise that the meaning and even the precise wording of the
Book of Odes and Classic of Filial Piety changed over time, their interpretation
of the images as illustrations necessarily stabilises the text and reinforces its ontological primacy.
And justifiably so. The aim of this inquiry is not to undermine the value of interpreting visual exegesis through reference to its written subject, or of putting what came first in history before what came after. This is in many instances an appropriate and necessary approach to exegetical art. My goal, rather, is to expose the train of thought that follows from the recognition of an artwork as exegesis, and thereby recognise the other interpretive possibilities that this categorisation forecloses. Even though art may appear in a conventionally exegetical format, this does not necessarily subordinate it to a textual precedent in the manner that the word ‘exegesis’ implies. Because the
4
Stephen Owen, Readings in Chinese Literary Thought (Cambridge: Council on East Asian Studies, Harvard University, 1992), p. 21.
5
Richard Barnhart, Li Kung-lin’s Classic of Filial Piety (New York: Metropolitan Museum of Art, 1993); Li Kung-lin’s Hsiao Ching T’u, Illustrations of the Classic of Filial Piety (Ph.D. diss., Princeton University, 1967).
6
Julia Murray, Ma Hezhi and the Illustration of the Book of Odes (Cambridge: University of Cambridge, 1993).
expressive potential of pictures is different from that of texts, they have in some cases the capacity to supplant the classic text as the primary locus of meaning and ontological basis for interpretation.
This process of pictorial self-canonising is evident in a Chinese book entitled Illustrations of the Three Ritual Classics (Sanlitu). Composed in the mid-tenth century, the Sanlitu is the earliest surviving example of a long tradition of illustrated exegesis on the Chinese ritual canon dating back to the second century CE. The ostensible subjects of this exegesis were the Three Ritual Classics (san li): The Record of Rites (Liji), the Rites of Zhou (Zhouli), and the Book of Ceremony and Etiquette (Yili).7 Principally compiled during the second half of the first millennium BCE (the exact dates are contested), these three texts were grouped together by the late Han exegete Zheng Xuan (127-200) as definitive records of the ceremonies and etiquette practiced by the aristocracy during the early Zhou dynasty (eleventh to eighth century BCE), which subsequent generations of Confucian scholars celebrated as an era of good governance and social harmony. The ‘illustrations of ritual’ (li tu) that Zheng Xuan produced as part of his commentaries on these texts are among the earliest recorded visualisations of the garments, sacrificial vessels, and other ceremonial paraphernalia named therein. They were followed by at least six other illustrated commentaries over the course of the first millennium CE. The earliest of these works to survive in full is the Sanlitu, which was compiled by the scholar Nie Chongyi and presented to the court of the first emperor of the Song dynasty, Zhao Kuangyin (r. 960-976), on 19 June 961. After review and minor modification, Nie’s illustrated text was disseminated in print as the
Newly Determined Illustrations of the Three Ritual Classics, with Compiled Commentaries (Xinding sanlitu jizhu).8 The earliest extant copy of this text is an 1175 reprint now housed at the National Library of China in Beijing.9
The twenty-fascicle text features over three hundred individual entries. Each entry begins with a relatively simple line-drawn depiction of an object, followed by textual citations from the ritual classics and earlier exegesis, as
7
For synopses of the three texts, including discussion of their historical transmission as well as important critical editions and translations, see Michael Loewe, ed. Early Chinese Texts: A Bibliographical Guide (Berkeley: Society for the Study of Early China, 1993), pp. 24-32, 234-243, 293-297.
8
Nie Chongyi’s biography in the fourteenth century History of the Song Dynasty (Songshi) is our primary source for the history of Nie’s recension of the Sanlitu. Tuotuo, et al., Songshi (Beijing: Zhonghua shuju, 1977), f. 431, pp. 12793-12797.
9
Nie Chongyi, Xinding sanlitu (Beijing: Zhonghua shuju, 1992 [Zhenjiang, 1175]). A critical edition is also available under the same title, punctuated and annotated by Ding Ding (Beijing: Qinghua daxue chubanshe, 2006).
well as additional comments by Nie Chongyi himself. The text describes, in order, the ceremonial hats, robes, and other garments for the imperial family, court nobility, and officials; the layout of ritual spaces; musical instruments; ceremonial archery equipment; banners and other processional insignia; jade insignia; jade implements used in sacrificial ceremonies; ritual vessels and other sacrificial implements; and mourning vestments and other funerary paraphernalia. The order of the chapters thus reveals a hierarchy that moves outward from the imperial person and downward from the most auspicious to the least auspicious rites.
In a postface, Nie Chongyi characterises the work as a comprehensive illustration of implements named in the Three Ritual Classics to be used as a reference for contemporary ceremony. The book thus purports dual purposes as both a descriptive commentary on classical texts and a prescriptive guide to the material aspects of contemporary etiquette. Whereas the text of Illustrations of
the Three Ritual Classics alternates between these two functions, the images
are almost exclusively prescriptive, displaying systematic disregard for the diversity of exegetical opinions expressed in the text. The usurpation of textual authority by the picture is exposed by this discrepancy between written and illustrated commentary.
Our first example is Nie’s entry on the so-called huang yi, or ‘yellow yi vessel’ [Figure 1]. The picture of this object shows a small handle-less cup with a short, slightly flared foot seated atop a much larger stand. Despite the simplicity of the monochrome rendering, the image conveys a considerable amount of information. It shows that the stand is composed of a foliated platter surmounting the tiered neck of a vase-shaped body with broad shoulders tapering to a flat base. A short gloss appended to the label identifies this stand as a zhou (more conventionally ‘saucer’), which the reader recognises as a discrete object from an illustration on the preceding page. The tiers of the neck are black, suggesting discrete decorative registers. At the center of the upper register is the top third of a circle, balanced on either side by pairs of roughly parallel lines. At the center of the lower register is a triangle with slightly concave sides and matching pairs of parallel lines immediately on either side. The base of the body bears five horizontal lines, above which can be seen a roughly triangular area framed by three nested parallel lines articulating five discrete lobes that echo the foliated edge of the platter above. This lobed pattern is reversed in single lines on either side, suggesting additional triangular registers on the unseen portions of the vessel. At the center of the triangular area, and on the otherwise unadorned cup above, are pairs of almond-shaped ellipses. We immediately recognise these as eyes on the basis of their shape, symmetrical arrangement, the discrete dots at their centers, and
the sockets suggested by the curved lines above and below. As the only explicitly representative element of the decorative scheme, these eyes attract the viewer’s attention and, together with its distinctive shape, present themselves as the most salient formal feature of the yellow yi.
Only a handful of the picture’s formal details are reflected in the accompanying text:
The yellow yi vessel holds fragrant liquor. The Officer of Zun and Yi Vessels states, “In the autumn and winter sacrifices, liquor is sprinkled on the ground using a jia-style yi vessel and a yellow yi vessel. Both have stands.” The king uses the jade ladle with a gui-shaped handle to make wine offerings to the sacrificial recipient and honor the gods. The queen uses the jade ladle with a zhang-shaped handle to make the subsequent offering. Zheng Xuan notes that “yellow yi vessel” means that it has yellow eyes, for which gold is used. The Suburban Animal Sacrifice says that “‘Yellow eyes’ refers to the superior zun vessel used for the vapor of fragrant liquor. ‘Yellow’ means ‘within.’ ‘Eyes’ means ‘that which makes the vapor clear.’ This means that when liquor is within, it is clearly visible on the outside.” This type of yi vessel and stand is entirely covered with golden lacquer.10
Nie Chongyi begins by introducing the function of the ‘yellow yi,’ citing the
Rites of Zhou as his authority. He then explains the respective types of ladle
that the king and his consort would use in conjunction with the yi. After this discussion of its function, he offers two different explanations for the vessel’s distinctive name. The first, from the Han exegete Zheng Xuan, is that ‘yellow’ refers to the yellow eyes (huang mu) that decorate the vessel’s surface. The second, an attenuated elaboration on the term ‘yellow eyes’ based on a reference from the Record of Rites, is that the exterior of the vessel somehow indicates whether or not there is liquor within. Passing no judgment on the relative merits of these differing interpretations, Nie closes with a comment on the lacquer used to decorate the yellow yi.
Comparison of text and image highlights the latter’s inversion of canon and commentary. While the illustration is ostensibly based on a classical source, the text reveals that this source authorises nothing more than the name ‘yellow yi.’ The varying interpretive possibilities enabled by the subsequent exegetical linking of this name with ‘yellow eyes’ show that even the most superficially obvious words – yellow and eyes – did not, within the flexible lexical framework of classical Chinese, offer sufficient cues for graphic representation. The text thus demonstrates that nothing about the name is
10
The Officer of Zun and Yi Vessels is a section in the Rites of Zhou; the Suburban Animal Sacrifice is a chapter in the Record of Rites.
formally explicit, making the picture less the visualisation of a textual description than an arbitrary image with a classical name. No authority is cited for the vast majority of the picture’s formal features.
The one physical characteristic that does appear to be based on a textual source – the eyes – actually advances the authority of visual comment over written canon. The decision to render eyes on the vessel’s surface represents an unequivocal rejection of the functional reading from the Record of Rites in favor of the Zheng Xuan’s formal literalism. This dismissal of a recognised classic for the opinion of a later exegete constitutes a radical departure from conventional exegetical practice. That this reversal happens only in picture is unsurprising. To have said it in word would have exposed its suspect logic. The graphic arrangement of the entry masks this inconsistency by subordinating the text to an image presented fait accompli.
The final line of the entry further appends text to image by shifting attention from the ritual classics to the image itself. By indicating that the yellow yi is entirely coated with gold lacquer, Nie offers supplementary formal information that the monochrome line drawing is ill-suited to communicate. The text offers no evidence that this interpretation of the vessel’s ‘yellow’ character is anything other than Nie’s own invention. This use of graphic infill – not color-by-number but color-with-letters – indicates that the real objects of exegesis are not the words of the ritual classics but the objects that they name. It belies the notion that Nie’s work was principally intended as a commentary on canonical texts.
What Illustrations of the Three Ritual Classics actually constitutes, rather, is a commentary on canonical pictures. The subordination of text to image graphically manifested on the page bearing the yellow yi is repeated throughout every other entry in the book. Illustrations scans as just that, a series of illustrations. What falls between these illustrations are written references to the objects drawn from the classics, their commentaries, and the insights of Nie himself [Figure 2]. This inter-graphic insertion of textual commentary mirrors the visual arrangement of the interlinear commentaries typical of the Chinese classics [Figure 3]. As a graphic membrane, the commentary in both instances frames and announces the canonical authority of the image-cum-classic.
The canonising of the graphic manifested in this interplay of text and image is contextually evidenced by the historical dissemination of Nie’s pictures. Records indicate that shortly after the book’s submission to court, the emperor commanded that its pictures be reproduced on the walls of the lecture
hall in the Directorate of Education.11 Located in the Song capital of Bianjing (modern Kaifeng), this institution was tasked with educating the most distinguished candidates for civil service positions from throughout the empire. The presentation of Nie’s images in this didactic setting, presumably shorn of the inter-graphic text, highlights their independent canonical status and reminds us that this status was ultimately the product of extra-textual fiat. As a self-contained subject of analysis, the format of Illustrations effectively canonises the images, but as a subject of history, it surrenders its agency to the imperial person.
The book also enunciates the scope of the emperor’s authority over these images. In the process, it exposes the hermeneutic that gave them shape. Contrary to the expectations of its title, this hermeneutic is not primarily directed toward the content of the ritual classics, but is based instead on an extra-textual logic of name-shape correspondence and pictorial representation. The most explicit illustration of this logic is the entry on the so-called xian-style zun vessel.
According to the Positions in the Radiant Hall, “The xian and xiang are zun vessels of the Zhou court.” The Officer of Zun and Yi Vessels states, “Two xian-style zun vessels are used for the morning stage of the spring and summer sacrifices. One holds black water, the other holds sweet liquor. The king uses a jade jue-cup to offer sweet liquor to the ritual recipient”. The Ritual Vessels relates, “The superior xi-style zun vessel of the court temple is placed to the west”. The commentary notes that the Rites of Zhou writes ‘xi’ as ‘xian’. In addition, the Mao Commentary on the Hymns in the Classic of Poetry says that this zun vessel is decorated with sand (sha) and feathers (yu). However, the commentaries of the [other] Mao and Zheng Xuan read the words ‘xian’ and ‘sha’ as having the same meaning as the ‘sha’ of ‘posha’ (dancing), and say that they all refer to zun vessels with images of phoenixes in a strutting posture on their surface. There is also the subcommentary to the Classic of Poetry by Wang Su, which says that “Rites calling for the xi and xiang-style zun vessels use zun made in the shape of a cow and elephant, with openings in their back.” Among the sacrificial vessels in use today are those shaped like a cow and elephant, with a lotus blossom seats carved on their backs. While these are not related to the zun vessels, they are very similar to those described by Wang. According to the illustration of Ruan [Chen], the xi-style zun vessel is decorated with a cow. He also notes that the lips of those used by the Noble Lords were decorated with ivory, while those used by the Son of Heaven were decorated with jade. His illustration shows a
11
zun vessel with an image of a cow painted on it, which is entirely different from what Wang Su describes. Taking this [description by Ruan] as a model feels acceptable. Here, I present it together with the illustrations by Zheng [Xuan]. Please select either.12
Nie Chongyi catalogs four distinct interpretations of the xian-style zun vessel, each based on different exegetes’ readings of the term xian. One interprets it as an indication that the vessel is feathered, another that it bears the image of a strutting phoenix, a third that it is shaped like a cow, and a fourth that it bears the image of the cow. In this case, Nie renders his judgment in both word and image, accepting the propositions that it refers to either a vessel adorned with a picture of a phoenix or a vessel adorned with a picture of a cow. He illustrates each of these two options and entreats the reader to “please select either” [Figure 4]. Given the fact that the book was compiled on imperial orders and presented to the emperor, it is safe to assume that the emperor was the intended subject of this entreaty.
The presentation of this choice exposes the insufficiency of Nie Chongyi’s hermeneutic as a tool for aligning the names of classical ritual objects with specific and singular forms. Although choices such as this are presented for only a handful of the implements in the book, their very presence gestures textually to the same imperative that the lecture hall murals denoted contextually – the canonisation of a set of ceremonial implements with specific shapes, dimensions, and décor. As Nie explains in his postface to Illustrations, this standard was meant to be timeless – both an embodiment of the normative forms of classical antiquity and a model appropriate for the present.13 When his hermeneutic failed to achieve this ideal, an arbitrary ruling was needed. The emperor supplied this.
Like the discrepancy between the paltry formal expectations of the text and comprehensive forms of the picture observed in the entry on the yellow yi, the interpretive doubt displayed in the entry on the zun vessel highlights the limited hold of the canonical text over the pictures. The appeal to the imperial arbitration heralds the fact that the image was less derived from the text than imposed upon it.
Although imperial arbitration invites capriciousness, the emperor’s whim is narrowly constrained. Nie Chongyi offers only two of the four available interpretations of the zun vessel for his consideration. The remainder he dismisses on expressly subjective grounds (literally translated, the original Chinese says that he “assessed them on the basis of human sentiment” [kui zhi
12
Nie Chongyi, Xinding sanlitu,14.4a. The Positions in the Radiant Hall is a chapter in the Record of Rites.
13
renqing]). Comparison of the two images to other zun vessels illustrated in the
book, such as the ‘mountain zun,’ suggests another more objective rationale [Figure 5]. All of the zun vessels have similar vase-like profiles and flared ring feet. In each case, the modifier, be it ‘xian’ or ‘mountain,’ is represented as a literal picture on the body of the vessel. In formal terms, the only difference between the two versions of the xian-style zun that Nie accepts is the subject of the motif. The other aspects of their design are all identical. This indicates a textually unstated but graphically unequivocal desire for formal consistency between objects of the same name. Applying feathers instead of a phoenix motif would have interrupted the word-to-picture translation of the adjective. Using a cow-shaped vessel instead of a vase with a picture of a cow would have disrupted the typological alignment of the word zun with vase-shaped vessels.
Driving Nie Chongyi’s formal decisions, then, is a pictorial logic. Although he consults the ritual classics and the visual suppositions of subsequent exegetes, his ultimate priority is to establish a coherent set of ritual implements according to the principle of name-form correspondence. If this means disregarding the opinion of a text, whether classic or commentary, then so be it. In the end, the logic of image-making overrides the logic of textual precedent.
These are but the most striking examples of a phenomenon that recurs throughout the Illustrations of the Three Ritual Classics. Nie Chongyi’s discussion of altar tables recognises minor formal variations between the tables of four different historical periods, but concludes that there is sufficient consistency for any to offer a viable model for the present. Here again he leaves the decision to the emperor, finding in the regularity of the form over time a justification that supersedes questions of historical precedence.14 Likewise, his dimensions for archery targets ignore the judgments of earlier exegetes in favor his own arbitrary measurements in “millet inches.”15 Time and again, the book willfully disregards classical precedent in order to erase interpretive uncertainty and thereby transform hypothetical pictures into canonical forms.
The canonising of image that occurs through the internal dynamics of
Illustrations of the Three Ritual Classics belies identification of the book as a
work of exegetic art. Although its title valorises liturgical texts that within the context of the Confucian tradition can arguably be understood as scripture, and its structure echoes the conventions of exegetic commentaries on the
14
Nie Chongyi, Xinding sanlitu, 13.7b-8a.
15
Confucian classics, its ultimate aim is not to comment on a classic but to canonise a new set of ritual implements. The book’s structural allusions to the established graphic and syntactic templates of exegetic literature should be understood, therefore, as the framing for a new exegetic object. The insertion of this graphic object into the exegetic template supplants the Three Ritual
Classics as the subject of exegesis, disintegrates these classics into fragmentary
citations, and works the fragments into the gilding for this frame. There it impresses upon the reader the erudition of Nie Chongyi and status of the images so framed. What could possess more authority that an object that shapes the most established canon to its contours?
This reading of Illustrations cautions against the easy identification of art as exegetic, and the subordination of image to text that this identification implies. As such, it invites reconsideration of the status of the images that one might situate into Hughes’s three categories of exegetical art. Imagining the experience of an illustrated book in visual terms – as a viewing instead of a reading – shifts the hermeneutic significance of the writing from the iconographic to the graphic. Whereas conventional art historiography post-Panofsky operates according to a dialectical tension between the foregrounding of the image as interpretive object and privileging of word as interpretive device, Illustrations of the Three Ritual Classics elicits interpretations that blur the boundary between word and image, setting both on a phenomenological continuum. These interpretations essentially erase the external object indexed by ‘exegesis,’ eliding the term’s epistemological distinctiveness and conflating it with other words of formal analysis (like color), which confer status unto objects through their extra-textual associations with socially-embedded hierarchies of value. When we consider the text in an illustrated handscroll or manuscript not as the object of illustration, but as a graphic endorsement of the illustration’s import, it becomes easier to understand why the pictures in
Illustrations enjoyed such a long afterlife.
Scholarship on the relationship between the normative models proposed by Nie Chongyi and the material culture of Confucian ritual in China has highlighted the extent to which Nie’s illustrations were criticized, first by eleventh century antiquarians who rejected Nie’s textual hermeneutic in favor of interpretations based on the epigraphic and formal analysis of antique Zhou dynasty vessels, then by the Emperor Huizong (r. 1100-1126), who commanded that they be stricken from the walls in the Directorate of Education and replaced with pictures based on objects in the court’s collection
of ancient ritual bronzes.16 And yet, as demonstrated by the ceramic historian Xie Mingliang, the illustrations remained in circulation as models for funerary implements until well into the fourteenth century.17 The book itself comes to us through an edition printed by the prefectural schoolmaster of Zhenjiang in 1175, more than fifty years after the imperial prohibition.18 The persistent valence of Illustrations undoubtedly owes a great deal to the imperfect dissemination of later officially sanctioned models and other historical coincidences, but it also implies that the book as a whole superseded the textual logic of its making. Although the hermeneutic of ritual reconstruction encoded in the text of Illustrations did not outlast the eleventh century, the text itself continued to circulate in conjunction with the pictures. This implies that its value extended beyond the linear logic of its words to the emotive force of their visual associations. The words endorsed the text through the import of their presence rather than the semantics of their meaning. In effect, they decorated the pictures, with all of the value distinctions that décor implies. By so eliding the boundary between image and word, Illustrations of the Three
Ritual Classics endured the disruptions of epistemic change.
16
Chen Fangmei, ‘Song guqiwuxue de xingqi yu Song fanggu tongqi,’ Meishushi yanjiu jikan, no. 10 (March 2001), pp. 37-160; Jeffrey Moser, Recasting Antiquity: Ancient Bronzes and Ritual Hermeneutics in the Song Dynasty (Ph.D. diss., Harvard University, 2010), pp. 232-239.
17
Xie Mingliang, ‘Beifang bufen diqu Yuanmu chutu taoqi de quyuxing guancha,’ Gugong xueshu jikan, vol.19, no.4 (2002), pp. 143-168.
18
Figure 2: Page with entries for the gui and fu vessels. Nie Chongyi, Xinding
Figure 3: Wei Liaoweng (1178-1237), Principle Meanings of the Book of
Ceremony and Etiquette (Yili yaoyi), printed 1252, f. 1, pp. 4b-6a. The three
lines of indented text headed with Chinese numbers in black boxes are from the Book of Ceremony and Etiquette. The remaining lines are all subsequent commentary.
Figure 4: Xian-style zun vessels. Nie Chongyi, Xinding sanlitu, 14.4a.
